This time last year the world was a different place. We had just published our Slow AI research, were coming off a failed Kickstarter campaign, and putting everything we had (and more) into planning the AIxDESIGN Festival. Fast forward to today: Slow AI is a thing.
But worldbuilding is an effortful exercise, with plenty of crashes along the way. With so much inertia supporting the status quo, its a gargantuan effort to build alternatives. We found that out the hard way.
TLDR: In 2025, design became about how to show up: to a space, a conversation, a movement. We improved our fundraising, and discovered that gardening new ideas requires a fallow season too.
TLDR: We threw our first IRL festival and published three chapters, three zines, and three artworks to reimagine AI, and kept building spaces for critical AI discourse. All while staying committed to making this work accessible, participatory, and actually fun.
We started 2025 by publishing the Slow AI Anthology: a series of zines, artworks, and non-linear chapters that compile our multi-method, decentralized, and participatory research on alternative narratives for AI.
This hot compost pile explores three pathways: Esoteric AI, Small AI, and Ancestral AI.
Esoteric AI challenges mainstream perceptions of AI as either disenchanted or enchanted, investigating the seeming binary of magic and technology. We examined the long history of predictive practices to learn how we might make modern predictive technologies, such as budgeting or health-tracking apps.
We asked: What can we learn from il/legitimate predictive technologies of the past?